A Path Planning Method Based on Basic RRT* Algorithm and Cagd-Based Curves Ffor Non-Holonomic Wheeled Mobile Robots

Eshtehardian, Ali | 2021

555 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 53752 (08)
  4. University: Sharif University of Technology
  5. Department: Mechanical Engineering
  6. Advisor(s): Khodaygan, Saeed
  7. Abstract:
  8. Throughout recent decades, one of the most important and challenging problems in robotics has been robot's path planning. Path planning means a robot must find its way from start to the goal point and track it without any collision to the obstacles. For this aim during the recent decades a wide variety of algorithms such as A*, Dijkstra and … have been proposed that some of which generate an optimal path as their output while others tend to create only a path regardless of its optimality. One of the most practical path planning methods is RRT algorithm which is executable in real-world applications, and by generating some nodes randomly then creating a tree-based graph, thereby outputting a path so as to be tracked by robot from start to goal. Output is not always optimum. Therefore, one can use RRT* instead of RRT for optimizing the path. On the other hand, since the path generated by this approach includes some discrete parts due to its tree-based entity, which cause getting non-holonomic robots into trouble (such robots require a smooth way based on the dynamic constraints), in this project the aim is to utilize one of the appropriate CAGD tools in order to smooth the incline of the RRT*'s output
  9. Keywords:
  10. Robot ; Nonholonomic Constraints ; Computer Aided Design (CAD) ; Geometric Design ; Path Planning ; Rapidly-Exploring Random Tree (RRT)Algorithm

 Digital Object List